Cost-effective control of 1080 bait-shy possums (Trichosurus vulpecula)

The brushtail possum (Trichosurus vulpecula) has been identified as a significant New Zealand conservation pest and a major wildlife reservoir of bovine tuberculosis (Tb; Mycobacterium bovis). To combat its continuing impact, central and local government agencies currently spend more than $50 million per annum on possum management activities. The current objective of this effort is to maintain possum population densities in selected areas below predetermined environmental and disease thresholds. Six toxicants are currently registered for possum control, with sodium monofiuoroacetate (1080) being the most extensively used. 1080 can be incorporated into various baits types and has been shown to be an extremely cost-effective method of initially removing <90% of a possum population. Unfortunately, 'bait shyness' induced among the surviving possums means that the efficacy of this acute-acting toxicant can decrease markedly when used repeatedly for maintenance control, which sometimes can be required at annual or biennial intervals. Given that there are currently few feasible alternative control options, 1080 bait shyness poses a threat to the sustainability of possum control in New Zealand. Three pen trial studies were conducted to investigate the best methods of preventing and mitigating 1080 cereal bait shyness. In each trial, 1080 cereal bait shyness was induced with an approximate LD₂₀ 1080 dose (0.8 mg.kg⁻¹ bodyweight). The first trial investigated the effectiveness of switching to alternative, slower-acting toxicants in cereal bait for maintenance control of 1080 cereal bait-shy possums. The three alternative toxicants tested in this trial (gliftor, cholecalciferol, marketed as Campaign® and brodifacoum, marketed as Talon®) had been identified in previous trials as being promising alternatives to 1080. The second trial investigated whether 1080 cereal bait-shy possums could be controlled using these alternative toxicants in an unfamiliar bait matrix (Pestoff® fruit paste). The third pen trial investigated the use of non-toxic cereal 'prefeed' and 'postfeed' as potential ways of inhibiting and overcoming 1080 cereal bait shyness. The postfeed result was also compared with changing to another unfamiliar bait matrix (Kiwicare fruit gel) containing 0.13% 1080. A bioeconomic possum model was then constructed to identify the most cost-effective control strategies to achieve a sustained 60% or 80% population reduction over a 10 year period, given that bait shyness can develop and that new alternative, slower-acting toxicants have become available for maintenance control. All of the alternative toxicants were relatively successful in the pens, killing 50-83% of the cereal 1080 bait-shy possums when in a familiar cereal bait. After the first night, consumption of the acute and subacute-acting toxicants decreased dramatically, whereas consumption of the chronic-acting brodifacoum toxicant increased progressively over subsequent nights. Some 1080 cereal bait-shy possums (40%) were killed when exposed to 0.8% 1080 in the unfamiliar paste bait. The effectiveness of the two alternative, sloweracting toxicants (cholecalciferol; a subacute-acting toxicant and brodifacoum; a chronicacting toxicant) was enhanced using the unfamiliar paste bait, with both achieving a 100% kill. Gliftor was not investigated further because of similarities between it and the existing acute-acting toxicant 1080. Prefeeding non-toxic cereal bait significantly reduced the number of possums that become 1080 cereal bait shy, with only 22% of the pre-fed possums developing an aversion to 1080 cereal bait compared to 97% of the non pre-fed groups. Postfeeding with non-toxic cereal bait, after the LD₂₀ 1080 dose, was relatively ineffective in reducing the number of 1080 bait-shy possum, with mortality of these possums being 30% compared with 0% of non post-fed possums. In contrast, the 0.13% 1080 gel bait killed 64% of the 1080 cereal bait-shy possums. The modelling simulation results suggest that it is possible to achieve a sustained 60% or 80% reduction in possum numbers using predominantly 1080-based control strategies, provided reasonably large (>100 ha) areas are controlled and 90% of the susceptible (i.e., non bait-shy) possums are killed in each operation. The 60% sustained population reduction can be achieved solely using 1080 control, however, the 80% population reduction will require the occasional use of an alternative, slower acting toxicant such as brodifacoum. Sensitivity analysis indicated that the most important variable influencing the overall success of these control strategies was the maximum rate of migration following control, which is influenced by control area size. With the high rates of migration that are sometimes observed into small (<100 ha) forest reserves, expensive permanent bait stations, containing brodifacoum bait, may be required in these to minimise the effects of immigration. The results of this study suggest that most 1080 cereal bait-shy possums will consume a lethal dose of a chronic brodifacoum toxicant provided their exposure to it is prolonged. However, subacute cholecalciferol poisoning symptoms do not appear to be delayed for long enough to be effective in the field when used in a familiar bait matrix. The effectiveness of 1080 and the alternative, slower-acting toxicants is enhanced when presented in the unfamiliar bait matrixes and these bait types should be field trialed in maintenance control operations. The low number of pre-fed possums (22%) that become 1080 bait shy following multiple doses of 1080 bait suggests that field managers should consider making greater use of non-toxic prefeed prior to bait station control operations. In conclusion, the modelling simulations suggest that sustained population reductions of 60-80% can be achieved using current control techniques. Further studies are required to determine the effectiveness of these strategies in the field. Possum researchers also need to investigate the factors determining population recovery in different sized control areas. The actual timing of control may vary between different control sites, and can only be established by direct measures of animal recovery and abundance.